Before losing itself in the sands of the 19th Century, the grand stream of Italian Renaissance architectural decoration made a last appearance in the Brumidi frescos of the Capitol Rotunda in Washington.

The artistic generation after Brumidi was trained in the Paris of that time to a more meticulous standard of execution, and tended to overlook greatness of conception where faults and weakness were easy to find.
But it is a great conception.
The open ceiling, with allegorical and classical figures thrown in masses against the sky: the closed frieze, formally divided into historical scenes and tightly tied to the stone walls, belong in their large ordering to the line of Correggio and his Baroque followers.
The descent may be remote, but this is surely the only full-scale example of that vigorous inheritance in the United States.

Constantino Brumidi designed the decorative scheme as a whole, in collaboration with the architect Charles U. Walter, at the time when plans were being made to replace the wooden dome of Bullfinch with the present much larger iron structure.
After many years and many interruptions he was able to finish the canopy fresco, and slightly less than half the frieze, beginning with the Liberty group opposite the East door, and ending with William Penn, all but one leg, when a tragic accident ended his career.
He left at his death sketches, drawn to scale, for the rest of the circle.
These were carried out not too faithfully by Filippo Costaggini, who began by supplying the missing member to the founder of Pennsylvania and noting in pencil, in Italian, that he "began at this point".

When Costaggini had used up all the sketches thirty-six feet of empty frieze were left over.
A blank undecorated void, plastered in roughcast, disfigured the wall of the Rotunda until 1951.
Then, advised by the Architect of the Capitol, the Joint Committee for the Library, traditionally responsible for the works of art in the building, ordered the space cleared and painted in fresco, to show "the Peace after the Civil War", "the Spanish-American War", and "the Birth of Aviation", to match as nearly as feasible Brumidi's technique and composition.
Later the cleaning and restoration were ordered, first of the older part of the frieze, finally of the canopy.
What follows is therefore a description of three separate undertakings, the new frescoing of the gap, and the successive essays in conservation, with some discussion of problems that arose in connection with each.

For the use of students and future restorers, a full, day-by-day record was kept of all three undertakings, complete technical reports on what we found and what we did.
These may be consulted in the office of the Architect of the Capitol, or the Library of Congress.

The first preliminary was inspecting the unfinished length of frieze, a jumble of roughcast and finish coats, all in bad condition.
It was decided to strip the whole area down to the bricks, and to replace the rough coats up to one inch thickness to agree with the older artists' preparation, with a mortar, one part slaked lime, three parts sand, to be put on in two layers.
Cartoons were drawn full size, after sketches had been made to satisfy all the authorities.
There was some difficulty here.
One had to manage the given subjects, three diverse recent events, so as to make them part of a classical frieze, -- that is, a pattern of large figures filling the space, with not much else, against a blank background.
Moreover, all three representations must be squeezed comfortably into little more than the length Brumidi allowed for each one of his.

When it was all arranged to fit, and not to interrupt the lengthwise flow of movement in the frieze, the cartoons were tried in place.
The scaffolding, a confusion of heavy beams hanging from the gallery above, was strong and safe, but obscured visibility.
Nothing could be seen from the floor, but by moving around the gallery one could get glimpses;
and we were able to decide on some amplification of scale.
To be sure of matching color as well as form, pieces of cartoon were traced on the roughcast, and large samples painted in fresco, then left two months to dry out to their final key.
Later it was gratifying to note that they had set so solidly as to be hard to remove when the time came.

The scaffold was the length of the space to be painted.
What bits of Brumidi and Costaggini could be reached at either end seemed in good order, though the roughish sandy surface was thick with dust.
Washed, they came out surprisingly clear and bright.
It could be seen that both artists used a very thick final coat of plaster, one half inch, and that both followed the traditional Italian fresco technique as described by Cennino Cennini in the 14th Century, and current in Italy to this day.
That is, they used opaque color throughout, getting solid highlights with active lime white.
Painting "a secco" is much in evidence.
A brown hatching reinforces and broadens shadows, and much of the background is solidly covered with a dark coat.
This brown is sometimes so rich in medium as to appear to be oil paint.

In our own practice, to have the last "intonaco" plaster coat thick enough to match, and at the same time to avoid fine cracks in drying, we found that it had to be put on in two layers, letting the first set awhile before applying the second.
The mortar was three parts sand to two of lime.
Some of the lime that is always on hand in the Capitol basement for plaster repairs was slaked several months for us;
but to make it stiffer, of a really putty-like consistency to avoid cracking, we added a little hydrated lime -- hard on the hands, but we could see no other disadvantage.
I am told that a mortar longer slaked might have remained longer in condition for painting.
As it was, it took the pigment well for six hours, enough for our purpose, and held it firmly in setting.
It was obvious that to match Brumidi, white must be mixed with all but the darkest tones.
Lime white, hard and brilliant, has a tendency to "jump" away from the other colors in drying, and also by its capacity to set, to preclude the use of ready-made gradations, so useful in decorative work.
In older Italian practice, lime, dried and reground "bianco sangiovanni", entered into such prepared shades.
For convenience we chose a stronger pigment, unknown to the early Italians or to Brumidi, titanium oxide, reserving the active lime white for highest lights, put on at the end of the day's stint.
Other pigments were mostly raw umber, some burnt umber, and a little yellow ochre.
This last was probably not in Brumidi's palette, but was needed to take the chill, bluish look off the new work next to the old, where softening effects of time were seen, even after thorough cleaning.
The use of "secco" we tried to restrict to covering joints.
Experience showed, however, that it is very difficult to paint a dark umber background in fresco that will not dry out spotty and uneven.
Later Brumidi and Costaggini will be seen coping with this same problem.
We were forced, as they were, to work a good deal of tempera into background and dark areas.
We made it by Doerner's recipe, five parts thoroughly washed cheese curd to one of lime putty;
ground together they made a strong adhesive, which became waterproof in drying.

Figure 1 was taken in 1953.
The new part is finished.
On the right is the Brumidi Liberty group, as it looked after cleaning operations, which had not yet come around to the other end;
where, of Costaggini, only some foliage has been washed, at the point where his work stopped.
One is led to speculate as to why the empty space was there, left for our century to finish.
Costaggini said it was Brumidi's fault in not providing enough material to fill the circle.
Brumidi's son later maintained that Costaggini had compressed and mutilated his father's designs, ambitiously coveting a bit he could claim for his very own.
This question might be settled by comparing the measurement of the actual circumference with the dimensions noted, presumably in Brumidi's hand, above the various sections of his long preparatory drawing, which has been kept.
Whosever fault, it is evident that Brumidi intended to fill out the whole frieze with his "histories" and come full circle with the scene of the discovery of California gold.
In painting a fresco, the handling of wet mortar compels one always to move from top to bottom and from left to right, not to spoil yesterday's work with today's plastering.
At the very first, then, Brumidi was required, by the classically pyramidal shape of his central group, to fill in the triangular space above the seated girl on Liberty's right, before starting on the allegorical figures themselves.
Here he put a small man, whose missing hands might have left his function doubtful, until comparison with the first sketches showed that when the artist came back to the beginning, this was to be the closing figure of the party of "forty-niners", and was to hold a basket.
One sees Costaggini's rendering of the same figure more than thirty feet away.
The photograph, Figure 1 of the completed frieze, shows how, having been separated from his fellows in useless isolation for eighty years, he has now been given a hand, and by juxtaposition (and the permission of the Committee), given a new job, to represent the witnesses of the first flight at Kitty Hawk in 1903.

The startlingly bright effect of the first washings led the Committee to order the rest of the Brumidi-Costaggini cycle cleaned and restored to go with them.
The fixed wooden scaffold was removed, and, so as to reach all the frieze, one of pipe, on wheels, built up from the floor.
Every few days, in the early morning, as the work progressed, twenty men would appear to push it ahead and to shift the plank foundation that distributed its weight widely on the Rotunda pavement, supported as it is by ancient brick vaulting.

On this giddy and oscillating platform over fifty feet from the floor, after a first dusting, we began to wash.
A most useful tool for wetting the surface without running down was made from a greenhouse "mist spray" nozzle welded to a hose connection, to be used at low water pressure.
A valve in the handle let us cut the pressure still lower.
One man sprayed, with a sponge in hand to check excess wetting.
A second assistant mopped with two sponges.
In parts a repeated sponging was needed, but everywhere we found that water alone was enough to restore the original brightness.
No soap or other cleaning agent was used that might bring in unwanted chemical reactions.
The painting "a fresco" stood up superbly;
a little of the "secco" came off.
Necessary retouching was put on at once.
Altogether we found the craftsmanship first rate, especially Brumidi's.
We were greatly helped by there being no traces of former restoring.
Apparently not more than dusting had ever been done, and not much of that.
The plaster was sound, the intonaco firmly attached all over, and the pigment solidly incorporated with it in all but a few unimportant places.

The greatest source of trouble was rain which had repeatedly flowed from openings above, soaking the surface and leaving streaks of dissolved lime, very conspicuous even after cleaning, particularly in the "Landing of Columbus", "Oglethorpe and the Indians", and "Yorktown".
Here the Architect, referring to the use of the Capitol as a public building, not a museum, requested some repainting to maintain decorative effect, rather than leaving blank, unsightly patches.

These frescos have had no care for eighty years.
With naked gas jets below and leaky windows above, enough to ruin wall paintings in any medium, they have survived, in a building long unheated in winter, hot and damp under the iron dome in summer.